I'm also wrong about the chimpanzees - recent research suggests that they share only 95% of genes with humans, not 98% - but at least I'm not off by a few orders of magnitude.
Read it, it doesn't state up to how many significant digits that percent is accurate so it doesn't prove I'm wrong. They may have just shorten it for brevity.
The convention in science is that you display as many digits as your measurement is significant to. If it were significant to two digits, it would've been 99%; if it were significant to four digits, it would've been 99.90%.
In any case, that number's backed up by several sources:
http://www.ornl.gov/sci/techresources/Human_Genome/project/i...
I'm also wrong about the chimpanzees - recent research suggests that they share only 95% of genes with humans, not 98% - but at least I'm not off by a few orders of magnitude.